Group Decision and Negotiation 14: 3–20, 2005.
C Springer 2005
Perceived Relative Power and its Influence
on Negotiations
REBECCA J. WOLFE
Princeton University, Department of Psychology, Green Hall, Princeton, NJ 08544, USA
KATHLEEN L. MCGINN
Graduate School of Business Administration, Harvard University, Boston, MA 02163, USA
Abstract
In an experimental study, we investigate perceived relative power in negotiations and its effect on the distribution of
resources and the integrativeness of agreements. We contrast perceived relative power with the objective individual
level measure of power often used in past research: the parties’ alternatives to a negotiated agreement. We found
that alternatives affected the distribution of outcomes, while perceived relative power and alternatives affected
the integrativeness of outcomes. We found that negotiating pairs who perceived a smaller difference in relative
power reached agreements of greater integrativeness than pairs who perceived a greater power difference, even
after controlling for alternatives and aspirations. We explore the implications of treating power in negotiations as
a perceived and relational construct.
Key words: alternatives, aspirations, negotiation, perceptions, power
Power imbalances are evident in most relationships, whether at the interpersonal level as
between a boss and her employee, at the inter-organizational level as between Microsoft and
the small start-ups it acquires, or at the international level as between the US and Mexico. The
power imbalance in these relationships stems from the asymmetry in dependence between
the parties, which contributes to an asymmetry in influence between the parties (Emerson
1962). For example, in an employment negotiation between a supervisor and a low-skill
applicant, the supervisor has a number of resources that the prospective employee needs,
such as money, benefits, vacation time, etc., and there are likely to be numerous applicants.
In contrast, the applicant has only one resource the employer needs, labor, and may have
few alternative opportunities for employment. The relative dependency of the applicant on
the employer may limit the information sharing necessary for finding integrative solutions,
and may result in the applicant accepting terms of employment favorable to the employer.
Rubin and Brown (1975, p. 260), in their seminal book on The Social Psychology of
Bargaining and Negotiation, assert that “mutual social influence . . . represents the fundamental strategic issue in bargaining”. To reach an agreement, each party needs to convince
the other to make a concession that he or she would not have made absent the influence of
the other. In an asymmetric relationship, in which the power balance between the parties
is unequal, the relatively high-power party is likely to have his or her interests addressed
during a negotiation, while the interests of the lower-power party may be ignored. This
4
WOLFE AND MCGINN
dynamic has implications not only for distribution of outcomes, which is likely to favor the
more powerful party, but also for integration. Without an exploration of the interests of both
sides, opportunities for creating value in unequal-power relationships may be lost. In this
study, we explore the effect of relative power within a negotiation on both the distribution
of resources and the integrativeness of outcomes.
Empirical evidence on the effects of power asymmetries on the integrativeness of negotiated agreements is mixed. Some studies confirm the above intuition that equal (relative
to unequal) power leads to greater integrativeness (Mannix and Neale 1993; McAlister,
Bazerman, and Fader 1986). Other studies find equal power is associated with relatively
lower integrativeness (Sondak and Bazerman 1991), and yet other research finds no effects
for power symmetry on integrativeness (Pinkley, Neale, and Bennett 1994). In all of this
research, power within a negotiation has been manipulated by varying either the value of the
best alternative or the number of alternatives. When power is manipulated through the value
of the best alternative, the sum of the parties’ alternatives is strongly associated with integrativeness. This result is not surprising – the parties need to surpass their alternatives in order
to reach agreement (Ben-Yoav and Pruitt 1984). Number of alternatives has also been shown
to have important effects on negotiated outcomes (McAlister, Bazerman, and Fader 1986),
and these effects are at least partially mediated by aspirations (Mannix and Neale 1993).
Although manipulations of power differed in the studies described above, in each of these
studies, the individual parties were aware of their own alternatives but not those of the other
party. These studies do not address how the level or number of one’s own alternatives might
influence a party’s perceptions of his or her power over the other, given the participants
have no information regarding the other’s alternatives. This critique does not invalidate the
basic findings of past research; parties’ alternatives no doubt have an effect on negotiation
outcomes. What is in question is whether the results from these previous studies are a
function of power – the dependence the parties feel towards one another – or a function of
individual-level anchors and aspirations stemming from the manipulated alternatives.
Power, as put forward by Emerson (1962), is driven by the difference across alternatives.
Following Emerson’s logic, the power of any one party cannot be assessed by considering
that party’s alternatives in isolation – it can only be assessed by looking at the relative
dependence of the parties on one another. Furthermore the comparison process that people
engage in when determining their power with regard to another person takes place at the
level of individual perceptions – people compare their dependence on their respective other
with what they believe is their respective other’s dependence on them (Blau 1964; Thibaut
and Kelly 1959). In the study presented here, we explicitly consider perceived relative power
and compare that to objective and individual-level resources. By examining the perceived
and relative nature of power, we hope to shed light on past inconsistencies in findings
regarding power and negotiation outcomes.
Defining power in negotiations
Within social psychology, views on the determinants of power fall into two camps: individualistic (Frost 1987; House 1988; McClelland 1975; Pfeffer 1992; Winter 1973) and
PERCEIVED RELATIVE POWER
5
relational (Blau 1964; Emerson 1962; French and Raven 1959; Kanter 1977; Kelman 1958;
Salancik and Pfeffer 1977). Individualistic notions of power focus on either a person’s
motivation to acquire power (House 1988; McClelland 1975; Winter 1973) or a person’s
individual traits, such as charisma, self-confidence and flexibility, which increase the likelihood of attaining positions of power (Frost 1987; Pfeffer 1992). Relational theories of
power examine the relative influence one party has over another. According to relational
theories of power, Party A has power over Party B to the extent that B is dependent on
A (Blau 1964; Emerson 1962) for goal achievement (Kelman 1958) or problem solving
(Kanter 1977; Salancik and Pfeffer 1977).
Relational theories of power vary with regard to whose perspective is considered – the
target of influence, the influencing agent, or both parties. French and Raven’s (1959) and
Kelman’s (1958) theories of power examine the power relationship from the perspective of
the target of influence: why a person is influenced by another. In contrast, Kanter (1977) and
Salancik and Pfeffer (1977) consider power from the perspective of the influencing agent:
how one’s situation within an organization or market affect the degree of influence one
has over others. Finally, definitions of power derived from social exchange theory examine
power from the perspective of both parties: how perceptions of relative dependency lead to
relative influence within the pair (Blau 1964; Emerson 1962).
In light of Rubin and Brown’s characterization of negotiation as fundamentally shaped
by mutual influence, a social exchange approach, which includes both parties’ perspectives,
appears to be the most appropriate foundation for studies of power in negotiations. According to social exchange theory, a party who is less dependent on her counterpart than her
counterpart is on her for an acceptable outcome has more power in the negotiation. Power
between the parties becomes more balanced as the parties become more equally dependent
on each other for some outcome.
Power, then, is determined by comparing A’s dependence on B to B’s dependence on
A. But dependency is only partially revealed in objective comparisons between the parties.
When the feature in question is socially and normatively determined – e.g., beauty, morality,
or power – people evaluate themselves through subjective social comparisons with focal
others (Festinger 1954; Thibaut and Kelley 1959; Ross and Nisbett 1991; Musseweiler
2003). When assessing one’s power within a negotiation, the other parties at the table are
the focal points of comparison; it is only relative to those others that the negotiator can
assess his or her own power within the bargaining setting. Social contextual variables,
past experiences, and expectations regarding future relationships affect the perceptions of
relative conditions (Anderson and Berdahl 2002; Thibaut and Kelley 1959). Assessing one’s
power accurately serves a critical social function and awareness of the distinctions from and
similarities with a counterpart allows people to navigate their social interactions smoothly
as they are better equipped to predict how someone will behave toward them and to respond
appropriately (Gill and Swann 2004).
Incorporating both perceptual and the relational features into the concept of power in
negotiations, we follow Emerson (1962) and Thibaut and Kelley (1959) in defining power as:
B has power over A to the extent that A perceives him-or herself as more dependent on
B than B perceives him-or herself as dependent on A.
6
WOLFE AND MCGINN
In our study, we contrast perceived relative power with three variables that have been
treated as manipulations or measures of power in past research: the objective, individuallevel resources each party brings to the negotiating table (alternatives); the perceptions of
those individual-level resources (reflected in aspirations); and the objective relative power
between the parties (differences across alternatives).
The role of perceived relative power in negotiations
Previous research on power in negotiations has focused largely on individual-level alternatives as the operationalization of power. Parties with objectively higher or more numerous
alternatives have been shown to achieve greater individual gains than those with objectively
lower or fewer alternatives (Mannix and Neale 1993). In addition, negotiations in which both
parties have highly-valued outside alternatives are likely to result in more integrative outcomes than negotiations in which one or both parties have less-valuable outside alternatives
(McAlister, Bazerman, and Fader 1986). But this work is largely uninformative regarding
the question of power balance, i.e. the equality or inequality of parties’ alternatives.
Aspirations have been proposed as the mechanism connecting alternatives to negotiated
outcomes. Alternatives seem to trigger goals, such that parties within a negotiation aim
to reach or exceed the value of the outside alternative. Mannix and Neale (1993) show
that giving a high aspiration to a lower-power party (the party with fewer alternatives)
in an unequal-power negotiation results in integrativeness that is equal to that produced
when both parties have many alternatives (equal-power condition). In this study, the authors independently manipulated aspirations and alternatives, though research suggests that
aspirations naturally derive from alternatives, such that higher alternatives lead to higher
aspirations (Pinkley, Neale, and Bennett 1994; Sondak and Bazerman 1991). Mannix and
Neale’s results add to the body of research illustrating that anchors, regardless of whether
they are presented as alternatives or aspirations, affect outcomes; higher anchors lead to
higher gains (Zetik and Stuhlmacher 2002).
While the research on alternatives and aspirations explores individual-level sources of
power, it provides little insight into the role of relative power. There has been an assumption
in the negotiation literature that power symmetry should lead to greater integrativeness
(Raiffa 1982; Zartman and Rubin 2000). Previous research examining this assumption
concludes that the sum of the parties’ alternatives affects integrativeness, but the relative
value of alternatives does not (Ben-Yoav and Pruitt 1984; Pinkley, Neale, and Bennett 1994;
Sondak and Bazerman 1991).
A process argument underlies the assumption that equality will lead to greater integrativeness: parties who share enough information to understand one another’s interests are more likely to find value creating trades than parties who either do not share
information or share information asymmetrically (O’Conner 1997; McGinn and Keros
2003). Integrative agreements result when parties communicate information regarding
their own interests and attend to the information the other party provides (Weingart,
Thompson, Bazerman, and Carroll 1990). Parties in equal power relationships are motivated to understand each other since they recognize that their own interests can be
met only if the other party’s interests are met too (Rubin, Pruitt and Kim 1994). In
PERCEIVED RELATIVE POWER
7
an unequal-power relationship, higher-power parties lack the motivation to try to understand lower-power parties (Fiske 1993; Keltner and Robinson 1997) and lower-power parties are reluctant to communicate their interests (Kipnis and Schmidt 1983; Snodgrass,
Hecht and Ploutz-Snyder, 1998). Consequently, unequal-power pairs are likely to be
at a disadvantage when trying to reach integrative agreements, relative to equal-power
pairs.
If power balance affects integrativeness of outcomes through the information exchange
within a negotiation, what might explain the inconsistent findings presented above? In previous research examining power asymmetries and negotiation integrativeness, it is unclear
how the parties perceived their relative power. Pinkley (1995) found that parties lacking
information about their opponents’ alternatives “filled in the blanks” and assumed that
their opponents had similar alternatives to themselves. These findings suggest that parties’
perceptions of the relative power they hold in a negotiation do not neatly correspond to
objective conditions. Parties may assume they are in an equal-power relationship regardless
of the objective difference between the alternatives.
Our understanding of the role of power in negotiations may be enhanced by disentangling the three distinct elements discussed above – alternatives, aspirations, and perceived
relative power – and distinguishing between the effects on distribution and the effects on
integrativeness. Our hypotheses draw from these distinctions. First, as shown in previous
research, alternatives are likely to affect both distribution and integrativeness.
Hla: As the quality of a party’s alternatives increase, individual payoffs are likely to
increase.
Hlb: As the quality of the parties’ alternatives increases, the integrativeness of the agreement is likely to increase.
These hypotheses reflect the effects of objective, individual-level resources on outcomes. Second, aspirations are also expected to affect both distribution and integrativeness:
H2a: As the aspirations of a party increase, individual payoffs are likely to increase.
H2b: As the joint aspirations of the parties increase, the integrativeness of the agreements
is likely to increase.
These hypotheses reflect the effects of the parties’ perceptions of their own individual
level resources on outcomes. Third, as argued above, we expect perceived relative
power to affect the integrativeness of agreements.
H3: As the perceived difference in resources increases, integrativeness of the agreements
is likely to decrease.
This hypothesis reflects the effects of perceived relative power on outcomes. Because
we conceive of our power variable at the relational level, we explore but do not hypothesize
the effects of this variable on individual payoffs. We also test for and discuss the distinctions between the perceptions reflected in aspirations, and those reflected in assessments of
relative power.
8
WOLFE AND MCGINN
Method
Participants and design
One hundred twenty-four undergraduates from five schools in the Northeast United States
participated in this experiment. Experimental sessions were run with 6 to 20 participants,
depending on attendance rates. Participants were paid $15 dollars for taking part in the
study, plus additional incentive pay based on the outcome of their negotiation. Within the
negotiation scenario used in this experiment, possible payoffs to the negotiators ranged
from $0 to $14,000 above their alternatives. In the experimental session, participants were
paid $1 for every $1,000 earned in the negotiation.
The negotiation simulated a second-round job interview between a candidate and a recruiter (adapted from Neale 1999). A 2 × 2 [Recruiter’s alternative: high (RH) or low
(RL) X Candidate’s alternative: high (CH) or low (CL)] between-subjects factorial design resulted in two cells in which power was objectively unequal between the parties,
RH/CL (n = 16) and RL/CH (n = 15), and two cells in which power was objectively
equal between the parties, RH/CH (n = 17) and RL/CL (n = 14). For both the Candidate
and the Recruiter, a high alternative was worth $9,500 and a low alternative was worth
$5,000 (above a baseline minimum). To control for possible status differences between
men and women, participants were assigned to same-sex dyads (Male dyads = 24; Female
dyads = 38).
To vary perceived relative power, participants were given information regarding their
own alternative and the alternative of their counterpart. In low-alternative conditions,
Recruiters/Candidates were told that their counterpart in the negotiation was their top
choice and that their alternative was worth $5,000. In the high-alternative conditions, Recruiters/Candidates were told that they had another attractive alternative (applicant/job offer)
worth $9,500 to them.
To manipulate knowledge of the other’s alternative, participants were given general
information about the other party and how the other party viewed them. Participants were
not given the exact value of their counterpart’s alternative. In the low other-alternative
conditions, participants were told that the other party was unhappy with their alternative
and thought of them as their top choice. Participants in the high other-alternative condition
were told that the other party had a very attractive alternative that would be difficult to beat.
Specific values for the counterparts’ alternatives were not provided to the participants to
heighten ecological validity, since people often have an idea of the worth of an opponent’s
alternatives, but rarely have complete, verifiable information about specific values placed
on those alternatives.
Materials and procedure
After participants arrived, they were asked whether they knew anyone else in the experimental session. To ensure that all dyads consisted of strangers, people who knew each other
were not assigned to the same dyad. Otherwise, participants were assigned randomly to a
partner, a role, and an alternative (own X other-alternative) condition.
PERCEIVED RELATIVE POWER
9
Participants had one-half hour to read and understand their roles and the issues under
negotiation. To settle, participants had to reach agreement on each of five issues: salary,
vacation time, bonus, insurance coverage, and moving expenses. The full scoring system
is presented in the Appendix (Table A1). Participants were told how they valued each of
the issues, but were not told their counterparts’ valuations. Salary was a continuous issue
(in dollars). The salary surplus available, i.e., the difference between the highest offer the
company could make and the lowest offer the recruit would even consider, was $8,000.
Possible outcomes on each of the other four issues were assigned monetary values, but
were limited to a small number of discrete choices. The two parties differed in the values they placed on the discrete issues, making integrative trades possible. The recruiter
valued vacation time and insurance coverage more than the candidate, and the candidate valued the bonus and moving expenses more than the recruiter. The four discrete
issues were worth $9,600 jointly if the parties compromised on each issue (i.e., settled
at the midpoint), but up to $14,400 jointly if the parties took advantage of all integrative
trades.
A fully integrative agreement, one that took advantage of all possible trades, resulted
in $22,400 ($8,000 salary + $14,400 across the discrete issues) available for distribution
between the parties. Because the bargaining zone is bounded by the parties’ alternatives, this
provided a fairly large bargaining zone ($12,400) when both parties had low alternatives,
a moderate bargaining zone ($7,900) when the alternatives were asymmetric, and a small
bargaining zone ($3,400) when both parties had high alternatives.
After reading their roles and asking any questions, but prior to negotiating, participants
completed questionnaires regarding their aspirations, their perceptions of power, and their
understanding of the materials and possible payoffs. (These measures are described in
detail below.) The experimenter then called each dyad one-by-one and led them to a private
room to negotiate. Once a pair finished the negotiation, they recorded their agreement on a
“contract” form included in the recruiter’s materials. The participants then returned to the
original room where they were debriefed and privately paid for their participation.
Measures
Individual alternatives were manipulated, as described above. The monetary values of the
alternatives were used in the individual-level analyses. The sum of the parties’ alternatives
was used as the dyadic measure, “joint alternatives.” Aspirations were measured through
the pre-negotiation questionnaire. Participants were asked: “How much do you hope to
achieve in this negotiation (i.e., your goal)?” The monetary value of the reported aspiration
was used in individual-level analyses. The sum of the reported aspirations was used as the
dyadic measure, “joint aspirations.”
The measure of perceptions of relative power was collected in the pre-negotiation questionnaire. Participants were asked to rate on a scale from 0 to 100, “What is your bargaining
power in this negotiation?” On the scale, 0 was labeled “my counterpart has all the power;”
50 was labeled “equal power;” and 100 was labeled “I have all the power.” The dyadic measure, “perceived relative power,” the operationalization of the degree of perceived power
10
WOLFE AND MCGINN
balance/imbalance in the pairs, was the absolute difference between the parties’ reported
perceptions of power.
We included three control variables in our dyadic level analyses. While the difference
in perceptions of power between ratings of, e.g., 70 and 90 are quantitatively equal to the
difference between ratings of, e.g., 20 and 40; there may be important qualitative differences
between negotiations in which both parties feel relatively powerful and those in which both
parties feel relatively powerless. Bachrach and Lawler (1981) highlight this distinction in
their theory of bargaining power, but do not provide clear predictions on how total power
affects integrativeness. To control for total power, we include the sum of the power ratings
in our dyadic analyses. To ensure that perceptions of relative power were not confounded
with differences in aspirations, the absolute value of the difference between the reported
aspirations, “difference in aspirations,” was also included as a control variable. To control
for objective differences in alternatives, we included a dummy variable, “objective relative
resources,” set to 0 when alternatives were equal (RH/CH or RL/CL) and 1 when the parties’
alternatives were unequal (RH/CL or RL/CH).
Outcome measures were based on the individual payoffs reported in the participants’
contract forms. The integrativeness measure was the sum of the participants’ individual
payoffs.1
Results
Four participants did not answer one or all of the pre-negotiation questions. These four
observations were excluded from the analyses, leaving a total of 120 individuals and 58
dyads for the following analyses. Means, standard deviations, and pairwise correlations
are presented in Tables 1 (individual-level variables) and 2 (dyadic-level variables). No
statistically significant main effects were found for sex, nor were any interactions with
sex statistically significant. The two statistically significant effects of role (candidate vs.
recruiter) are noted below. All other analyses were collapsed across sex and role.
Table 1. Means and pair wise correlations of individual-level variables
Mean
(SD)
Own alternativea
7250 (2260.30)
Other’s alternativea
7209.09 (2259.92)
Aspirations
10672.73 (3149.69)
Own
alternative
Other’s
alternative
Aspirations
Perceived
power
Individual
gains
1.00
−0.04
1.00
0.46∗∗,b
∗∗
0.02
Perceived power
60.73 (15.29)
0.27
−0.27
Individual gains
10085 (1746.26)
0.38∗∗
−0.13
1.00
∗∗
0.13
1.00
0.31∗∗
0.15
1.00
N = 110 [10 (5 dyads) impasses are omitted.]
Manipulated variable ($5,000 or $9,500).
b
The correlation between aspiration and own alternative varied significantly by role. Recruiters, r = 0.65, p <
0.0001; Candidates, r = 0.30, p < 0.05.
∗
p ≤ 0.05.
∗∗
p ≤ 0.01.
a
11
PERCEIVED RELATIVE POWER
Table 2. Means and pair wise correlations of dyadic-level variables
Mean
(SD)
Objective Perceived Difference
Joint
Joint
relative relative in
Total Joint
alternatives aspirations resources power
aspirations power gains
Joint alternatives
14500 (3181.98)
1.00
Joint aspirations
21527.59 (4653.87)
0.44∗∗
1.00
0.51 (0.50)
0.00
−0.11
1.00
Perceived relative 16.51 (13.54)
power
−0.04
−0.11
0.25∗∗∗
Difference in
aspirations
2889.66 (2802.67) −0.02
0.22
−0.01
0.05
1.00
Total perceived
power
121.23 (22.10)
0.01
−0.00
0.00
−0.08
−0.24
1.00
Joint gains
20283.96 (1595.36)
0.37∗∗
0.03
−0.30∗
0.05
0.18
Objective relative
resources
(equal = 0)
0.34∗
1.00
1.00
N = 53 (5 impasses are omitted).
∗
p ≤ 0.05.
∗∗
p ≤ 0.01.
∗∗∗
p ≤ 0.07.
Manipulation check
To assure that the participants understood their instructions, they were asked to state the
monetary value of their own alternative. Results provide evidence that the instructions
regarding alternatives were clearly understood. Participants in the high-alternative condition
stated a significantly higher alternative than participants in the low-alternative condition
[High alternative, M = 9443.75, SD = 337.06 (N = 64); Low alternative, M = 5261.81,
SD = 1344.88 (N = 55); one-tailed t(117) = −24.03, p < 0.0005].2
Individual-level outcomes
We turn first to our hypotheses regarding the effects of objective and perceived resources on
the distribution of surplus. We examined how alternatives, aspirations, and perceptions of
power affected individual payoffs. Five dyads did not reach agreement and were removed
from these analyses.3 In an OLS regression testing the combined effects of alternatives,
aspirations, and perceptions of power on individual payoffs, controlling for other’s alternative, we found that the effects of aspirations and alternatives were moderated by the role of
the participant. We therefore present the regressions for candidates and recruiters separately
in Table 3. Perceptions of power at the individual level had no effect on individual payoffs
in either role. For candidates, own alternative was a significant predictor of individual gains
– as one’s own alternative increased, individual gains increased. For recruiters, aspirations
were a significant predictor of individual gains – as aspirations increased, individual gains
increased. These findings provide partial support for Hla and H2a.
12
WOLFE AND MCGINN
Table 3. OLS regression of perceived and objective individual
resources on individual outcomes for candidates and recruiters
Candidates only
Own alternative
Other’s alternative
Aspirations
Perceptions of power
Standard coefficient (SE)
0.41∗∗ (0.11)
−0.20 (0.11)
0.05 (0.07)
−0.14 (20.26)
N = 53
F (4, 48) = 2.64∗
Adj.R2 = 0.11
Recruiters only
Own alternative
Other’s alternative
Standard coefficient (SE)
0.01 (0.13)
−0.09 (0.10)
Aspirations
0.47∗∗ (0.09)
Perceptions of power
0.10 (12.58)
N = 53
F (4, 48) = 4.30∗∗
Adj. R2 = 0.20
∗
p ≤0 .05.
p ≤ 0.01.
∗∗
In further post hoc analyses, we found that aspirations mediate the effects of alternatives
for recruiters.4 As noted in Table 1, the correlation between aspiration and own alternative
varied significantly by role. Though the two are significantly correlated for both roles, the
relationship is much stronger for recruiters (r = .30 for candidates; r = .65 for recruiters).
Testing for mediation in the recruiter’s role (Baron and Kenny, 1986), when recruiter’s payoff
is regressed on recruiter’s alternative, candidate’s alternative, and recruiter’s perception of
power, the beta for recruiter’s alternative is significant (β = 0.32, p < .05), though the full
regression is only marginally significant (F (3, 49) = 2.51, p < 0.10; Adj. R2 = 0.08).
When recruiter’s aspiration is added simultaneously with the other independent variables,
as shown in Table 3, the recruiter’s alternative coefficient is no longer significant, while the
coefficient for aspirations is significant and the full regression reaches significance. This
provides evidence that aspirations are fully mediating the effects of alternatives on payoffs
for those in the recruiter role.
Dyadic level outcomes
We present the OLS regression testing the hypotheses related to integrativeness of the
agreements in Table 4. Confirming previous studies and supporting H1b, integrativeness increased as the sum of the alternatives increased, though this effect was marginally significant
(p = 0.06). We found an important caveat to the conclusion that integrativeness increases
13
PERCEIVED RELATIVE POWER
Table 4. OLS regression of perceived and objective individual and relative
resources on integrativeness
Standard coefficicnt (SE)
Joint alternatives
0.27∗∗∗ (0.07)
Joint aspirations
0.19 (0.04)
Objective relative resources (equal = 0)
Perceived relative power
0.13 (407.39)
−0.29 (15.23)
Differences in aspirations
0.07 (0.07)
Total perceived power
0.17 (9.25)
N = 53
F(6, 46) = 3.15∗∗
Adj. R2 = 0.20
∗
p ≤ 0.05.
p ≤ 0.01.
∗∗∗
p ≤ 0.06.
∗∗
with the sum of the parties’ alternatives. Of the five dyads that did not reach agreement
(92.4% agreement rate), four of the dyads were in the HH condition and one was in the
HL condition. When impasses were included in the overall regression of integrativeness
(using sum of the alternatives as the integrativeness measure for impasses), the sum of
the alternatives no longer significantly predicted integrativeness.5 A logistic regression of
agreement (0 = impasse; 1 = agreement) on joint alternatives, joint aspirations, objective
relative resources, perceived relative power, total perceived power and difference in aspirations revealed the reason for this – a significant negative effect for joint alternatives (p <
.005), as well as a significant negative effect for objective relative resources ( p < .005),
though the overall regression was not statistically significant (χ 2 = 9.67, p = .14). While
higher joint alternatives increase the value of the agreements reached, they also increase the
likelihood of impasse. The perceptual variables, joint aspirations, perceived relative power,
total perceived power and difference in aspirations did not significantly affect agreement
rates, nor were their effects on integrativeness altered when impasses were included in the
analyses.
The effect of joint aspirations on integrativeness did not remain significant when controlling for joint alternatives, objective relative resources, perceived relative power, total
perceived power and differences in aspirations. Thus, H2b was not supported.
Our hypotheses regarding the negative effect for differences in perceived relative power
on integrativeness (H3) was supported. As differences between the parties’ perceptions of
power increased, integrativeness decreased, even after controlling for the other predictors.
We offer additional analyses exploring the difference between aspirations and perceived
relative power. Using an Analysis of Variance, we tested for effects of both own and other’s
alternatives on aspirations and perceived power. Reported aspirations were significantly
affected by one’s own alternatives [High alternative, M = 12050, SD = 2554.86 (N = 64);
Low alternative, M = 9241.07, SD = 2876.66 (N = 56), F(1, 112) = 33.38, p < .005], but
not by the counterpart’s alternative [High other’s alternative, M = 10820.64, SD = 3217.70
14
WOLFE AND MCGINN
Figure 1. Mean Aspirations by Own and Other’s Alternative.
(N = 63); Low other’s alternative, M = 10649.12, SD = 2863.57 (N = 57), F(1, 112) = 0.04,
p = ns]. There was no statistically significant interaction between the parties’ alternatives
on aspirations.6 These results are presented graphically in Figure 1.
In contrast, both own and other’s alternatives influenced perceptions of power (Own
alternative: F (1, 112) = 8.73, p < .005; Other alternative: F (1, 112) = 7.39, p < .01).
Parties in a high-alternative condition paired with a participant in a low-alternative condition
reported the highest perceived power [M = 68.21, SD = 13.42 (N = 28)]; parties in dyads
with equal-value alternatives reported their power to be roughly equal [M = 60.75, SD
= 50.07 (N = 60)]; and parties in a low-alternative condition paired with a participant in
a high-alternative condition reported the lowest levels of power [M = 52.69, SD = 14.75
(N = 28)]. There was no statistically significant interaction between the parties’ alternatives
on perceptions of power. These results are presented graphically in Figure 2.
Figure 2. Mean Perceptions of Power by Own and Other’s Alternative.
PERCEIVED RELATIVE POWER
15
Discussion
We have provided a more complete conceptualization of the effects of power on negotiation
outcomes by disentangling perceptions of relative dependency – power – from objective
resources – alternatives. Negotiators rated themselves as more or less powerful based not
only on their own alternative, but also the alternatives of their counterparts. Furthermore,
alternatives and perceived relative power have different effects on negotiation outcomes.
Individual-level resources, and the aspirations which reflect the perceptions of those resources, drive individual-level payoffs. Perceptions of relative power, however, have no
significant effects on the distribution of payoffs. In contrast, the effects of perceived power
are found in the integrativeness of the outcome. Even after controlling for the sum and
difference in objective resources across the parties, the sum and difference in aspirations,
and total perceptions of power, perceptions of relative power had predictable effects on joint
outcomes: those pairs in which both parties’ perceptions of power were roughly equal were
able to achieve significantly greater integrativeness than those pairs in which there were
greater differences in perceived power.
How people view their relationship, whether as one between relatively equal- or unequalpower parties, affects their motivation for negotiating with one another and subsequently,
their behavior. When people recognize their mutual dependence on one another, they are
more likely to recognize that helping their counterparts achieve their goals, in turn, will
advance their own interests (Rubin, Pruitt and Kim 1994). This instrumental concern for
others as well as oneself has been associated with greater integrativeness (Ben-Yoav and
Pruitt 1983; Rubin, Pruitt and Kim 1994). When people perceive themselves in an equalpower relationship, they are motivated to find trades that increase the value of the deal,
which allows both parties’ interests to be satisfied.
Past research has proposed that power works through aspirations (Mannix and Neale
1993; Pinkley, Neale, and Bennett 1994; Sondak and Bazerman 1991). Our findings suggest
that, while the effects of alternatives may sometimes work through aspirations, aspirations
and perceptions of power are conceptually and empirically distinct constructs. Participants’
perceptions of their power in a negotiation were unrelated to their aspirations in that same
negotiation, and the two variables derived from different sources and had separate effects on
negotiated outcomes. Specifically, perceptions of power incorporated both own and other’s
alternatives and had a strong effect on the integrativeness of the outcomes. In contrast,
reported aspirations reflected only one’s own alternative – participants ignored their counterparts’ resources when setting their goals for the negotiation. Consequently, aspirations
affected integrativeness in the same manner as alternatives – as joint aspirations increased,
integrativeness increased, but this effect was not significant after controlling for the parties’
alternatives. Moreover, we find that aspirations, but not perceptions of relative power, are
related to payoffs at the individual level.
At first glance, considering the other’s alternatives when determining one’s power
but not doing so when determining one’s aspirations may appear to be a cognitive error – the less power you perceive yourself to have in a negotiation, the less you should
hope to achieve. Such an error is consistent with a growing body of research illustrating
ways in which individuals ignore information about others in decisions and negotiations
16
WOLFE AND MCGINN
(Camerer and LaVallo 1999; Moore in press; Ball, Bazerman, and Carroll 1991). On a
strategic level, however, there may be a defensible rationale for paying attention to relative
resources when calculating power in a negotiation, but ignoring relative resources when calculating aspirations. In order to predict and be ready to respond to how one’s counterpart is
going to behave, it is important to accurately understand the power balance in a negotiation.
But, since higher aspirations lead to higher payoffs (Zetik and Stuhlmacher 2002), it may be
advantageous to have high aspirations even when in a low power position. Recent research
by Pittutla and Murnighan (2003) provides additional support for the disconnect between
perceptions of power and aspirations. In a series of experiments using ultimatum and dictator games, the authors found that when people lost power, people’s perceptions of power
reflected this shift, but their demands did not; their allocations remained the same. Future
research is needed to provide a deeper understanding of how negotiators develop beliefs
about power and aspirations, and how these interact with one another to affect negotiation
processes and outcomes.
Our results revealed unexpected role (candidate vs recruiter) differences in the effects
of alternatives and aspirations on individual payoffs. For recruiters, aspirations were highly
correlated with alternatives (r = 0.65) and fully mediated the effects of alternatives on individual payoffs; for candidates, there was less of a connection between alternatives and aspirations (r = 0.30), and aspirations had no effect on payoffs. Past research has shown that the
context of the negotiation and the roles people play affects the way participants interpret the
information they are given (Neale, Huber and Northcraft 1987), although later research suggested these differences are often due to an asymmetric presentation of alternatives in most
experimental materials (Ritov 1996). We suggest a more social explanation-participants
may be bringing in their “real world” assumptions about the roles. Here, a candidate was
looking for a job and a recruiter was trying to fill a position, giving each a different perspective on risk in the negotiation. Recruiters may have felt it was reasonable to be “pushy”
given a high aspiration, while candidates may feel somewhat at the mercy of a recruiter,
regardless of one’s aspiration level. These social perceptions of the role may have had an
implicit effect on their behavior (Bargh et al. 1995; Galinsky, Gruenfeld, and Magee 2003).
Our results provide further evidence that people consider more than the economic structure of a negotiation when determining their power. Previous research has shown that people
also consider their roles (Olekalns 1991), their individual values (Coleman 1998), their own
and others’ personalities (Anderson and Berdahl 2002; Georgesen and Harris 2000), and the
time horizon for the relationship with the respective other (Larson 1992) when determining
their power. Our research also raises, but does not address, questions about the effect of
different sources of power on people’s evaluation of the power relationship. In this study, we
only examine one source of power-alternatives – but as French and Raven (1959) show, there
are multiple sources of power. Some sources of power may weigh more heavily than others
in determining a particular power relationship. For example, asymmetric coercive power
may be more detrimental to integrativeness in a negotiation than power asymmetries based
on held resources or information. Further research is needed to uncover the similarities and
differences stemming from different sources of power.
The main purpose of this study was to explore how the perceived balance of power in
negotiations affects outcomes. Our results support our hypothesis that as perceived equality
17
PERCEIVED RELATIVE POWER
increases, integrativeness increases. But we did not investigate the process through which
parties who perceive their relationships as more equal reach agreements of greater integrativeness. The effects of relative power on negotiations appear to work through dyadic-level
mechanisms, such as the information sharing between the parties, rather than individuallevel mechanisms, such as goal setting or anchoring. This is consistent with a growing body
of research examining the effects of power differences on behavior (See Keltner, Gruenfeld,
and Anderson 2003, for a review). For example, parties in equal-power relationships are
more likely to reciprocate each other’s actions (Buunk et al. 1993; Lawler and Yoon 1993)
and are more likely to ask information-seeking questions of one another than unequal-power
parties (Giebels, De Dreu, and van de Vliert 1998), behaviors that are beneficial in negotiations. Future research could add important insights into our understanding of power in
negotiations by investigating how the process of negotiation varies across parties in equaland unequal-power relationships.
In conclusion, objective, structural variables, such as the resources the parties bring to
the negotiation, play an important role in determining the agreements the parties reach. But
these factors tell only part of the story. Resources reflect what each party hopes to achieve in
the negotiation, as previous research has shown, but resources also influence the perceived
power symmetry across the parties. When perceiving the relationship as more equal, parties
are likely to reach agreements of greater integrativeness–advantages lost when viewing the
relationship as asymmetric. By examining power as a perceived, relational variable, we can
gain a fuller understanding of the role of power in negotiations.
Appendix
Table A1. Payoff schedule
Issue
Options
Candidate
Payoffs ($)
Bonus
10%
4000
0
8%
3000
400
6%
2000
800
4%
1000
1200
2%
0
1600
Vacation time
Moving expenses coverage
Recruiter
Payoffs ($)
25 Days
1600
0
20 Days
1200
1000
15 Days
800
2000
10 Days
400
3000
5 Days
0
4000
100%
3200
0
90%
2400
200
80%
1600
400
(Continued on next page)
18
WOLFE AND MCGINN
Table A1. (Continued )
Issue
Insurance coverage
Salary($)*
Options
Candidate
Payoffs ($)
Recruiter
Payoffs ($)
70%
800
600
60%
0
800
Plan A
800
0
Plan B
600
800
Plan C
400
1600
Plan D
200
2400
Plan E
0
3200
50,000
8000
0
48,000
6000
2000
46,000
4000
4000
44,000
2000
6000
42,000
0
8000
∗
These are just examples of possible salaries. You may agree to any salary between $42,000
and $50,00.
Notes
1. We use joint profit as a proxy for integrativeness for the following reasons: 1) It is the most straight forward
and intuitive measure of integrativeness, 2) according to Tripp and Sondak (1992), if a relevant portion of the
pareto frontier has a slope of -1, there is little risk that using joint profit would lead to a different conclusion
than using an integrativeness measure of efficiency and 3) De Dreu and Carnevale (2002), state that although
there are differences between the measures in theory, empirically the measures have been strongly correlated.
2. One participant did not complete the manipulation check question.
3. The reported findings remained significant when impasses were included in the analyses.
4. It is clear that aspirations are not mediating the effects of alternatives for candidates, since the coefficient for
aspirations is not significant in the full regression on candidate payoffs.
5. Otherwise, there were no significant differences in the analyses when impasses were included.
6. In addition to the main effect of alternative, there was a main effect of role, such that candidates had higher
aspirations than recruiters [Candidates, M = 11208.33, SD = 3121.40 (N = 60), Recruiters: M = 10270,
SD = 2913.28 (N = 60), F(1, 112) = 5.02, p = .02]. The main effects of alternative and role were qualified
by a role X alternative interaction, such that among participants in the low-alternative condition, recruiters
had lower aspirations than candidates [RL: M = 8174.07, SD = 2678.20 (N = 27); CL: M = 10234.48,
SD = 2733.95 (N = 29); two-sided t-test (54) = 2.85, p = .006).
References
Alinsky, S. D. (1971). Rules for Radicals: A Practical Primer for Realistic Radicals, New York: Vintage Books.
Anderson, C. and J. Berdahl. (2002). “The Experience of Power: Examining the Effects of Power on Approach
and Inhibition Tendencies,” Journal of Personality and Social Psychology 83, 1362–1377.
Bachrach, S. B. and E. J. Lawler. (1981). Bargaining: Power, Tactics and Outcomes, San Francisco: Jossey-Bass.
Ball, S. B., M. H. Bazerman, and J. S. Carroll. (1991). “An Evaluation of Learning in the Bilateral Winner’s Curse,”
Organizational Behavior and Human Decision Processes 48, 1–22.
PERCEIVED RELATIVE POWER
19
Bargh, J. A., P. Raymond, J. B. Pryor, and F. Strack. (1995). “Attractiveness of the Underling: An Automatic
Power Sex Association and its Consequences for Sexual Harassment and Aggression,” Journal of Personality
and Social Psychology 68, 768–781.
Barron, R. M. and D. A. Kenny. (1986). “The Moderator-Mediator Variable Distinction in Social Psychological
Research: Conceptual, Strategic, and Statistical Considerations,” Journal of Personality and Social Psychology
51, 1173–1182.
Ben-Yoav, O. and D. G. Pruitt. (1984). “Resistance to Yielding and the Expectation of Cooperative Future Interaction in Negotiation,” Journal of Experimental Social Psychology 20, 323–335.
Blau, P. M. (1964). Exchange and Power in Social Life, New York: Wiley.
Buunk, B. P., B. J. Doosje, L. G. J. M. Jans, and L. E. M. Hopstaken. (1993). “Perceived Reciprocity, Social
Support, and Stress at Work: The Role of Exchange and Communal Orientation,” Journal of Personality and
Social Psychology 65, 801–811.
Camerer, C. and D. Lovallo. (1999). “Overconfidence and Excess Entry: An Experimental Approach,” The American Economic Review 89, 306–319.
Coleman, P. T. (1998). “Implicit Theory and Situational Priming Effects on Power Sharing Decisions: An Experimental Study, Working Paper,” Columbia University.
De Dreu, C. K. W. and P. J. Carnevale. (2003). “Motivational Bases of Information Processing and Strategy in
Negotiation and Social Conflict,” in M. P. Zanna (ed.), Advances in Experimental Social Psychology, Vol. 35,
New York: Academic Press.
Emerson, R. M. (1962). “Power-Dependence Relations,” American Sociological Review 27, 31–40.
Festinger, L. (1954). “A Theory of Social Comparison Processes,” Human Relations 7, 117–140.
Fiske, S. T. (1993). “Controlling Other People,” American Psychologist 48, 621–628.
French, J. R. P. and B. H. Raven. (1959). “The Basis of Social Power,” in D. Cartwright (ed.), Studies in Social
Power, Ann Arbor: University of Michigan Press.
Frost, P. J. (1987). “Power, Politics and Influence,” in F. M. Jablin (Ed.), Handbook of Organizational Communication. Newbury Park, CA: Sage.
Galinsky, A.D., D. H. Gruenfeld, and J. C. Magee. (2003). “From Power to Action,” Journal of Personality and
Social Psychology 85, 453–466.
Georgesen, J. C. and M. J. Harris. (2000).“The Balance of Power: Interpersonal Consequences of Differential
Power and Expectancies,” Personality and Social Psychology Bulletin 26, 1239–1257.
Giebels, E., C. K. W. De Dreu, and E. van de Vliert. (1998). “The Alternative Negotiator as the Invisible
Third at the Table: The Impact of Potency Information,” International Journal of Conflict Management 9,
5–21.
Gill, M. J. and W. B. Swann. (2004). “On What it Means to Know Someone: A Matter of Pragmatics”, Journal of
Personality and Social Psychology 3, 405–418.
House, R. J. (1988). “Power and Personality in Complex Organizations,” Research in Organizational Behavior
10, 305–357.
Kanter, R. M. (1977). Men and Women of the Corporation, New York: Basic Books, Inc.
Kelman, H. C. (1958). “Compliance, Identification, and Internalization: Three Processes of Attitude Change,”
Journal of Conflict Resolution 2, 51–60.
Keltner, D. K., D. H. Gruenfeld, and C. Anderson. (2003). “Power, Approach and Inhibition,” Psychological
Review 110, 265–284.
Keltner, D. K. and R. J. Robinson. (1997). “Defending the Status Quo: Power and Bias in Social Conflict,”
Personality and Social Psychology Bulletin 10, 1066–1077.
Kipnis, D. and S. M. Schmidt. (1983). “An Influence Perspective on Bargaining within Organizations, in M. H.
Bazerman and R. J. Lewicki (Eds.), Negotiating in Organizations, Beverly Hills: Sage, pp. 303–319.
Larson, A. (1992). “Network Dyads in Entreprenurial Settings: A Study of the Governance of Exchange Relationships,” Administrative Science Quarterly 37, 76–105.
Lawler, E. J. and J. Yoon. (1993). “Power and the Emergence of Commitment Behavior in Negotiated Exchange,”
American Sociological Review 58, 465–481.
Mannix, E. A. and M. A. Neale. (1993). “Power Imbalance and the Pattern of Exchange in Dyadic Negotiation,”
Group Decision and Negotiation 2, 119–133.
20
WOLFE AND MCGINN
McAlister, L., M. H. Bazerman, and P. Fader. (1986). “Power and Goal Setting in Channel Negotiations,” Journal
of Marketing Research 23, 228–236.
McClelland, D. M. (1975). Power: The Inner Experience, New York: Irvington Press.
McGinn, K. L. and A. T. Keros. (2002). “Improvisation and the Logic of Exchange in Socially Embedded Transactions,” Administrative Science Quarterly 47, 442–473.
Moore, D. A. (in press). “The Unexpected Benefits of Final Deadlines in Negotiation,” Journal of Experimental
Social Psychology.
Musseweiler, T. (2003). “Comparison Processes in Social Judgment: Mechanisms and Consequences,” Psychological Review 110, 472–489.
Neale, M. A. (1999). New Recruit. Evanston, IL: Dispute Resolution Research Center, Kellogg Graduate School
of Management, Northwestern University.
Neale, M. A., V. Huber, and G. Northcraft. (1997). “The Framing of Negotiations: Contextual Versus Task Frames,”
Organizational-Behavior-and-Human-Decision-Processes 39, 228–241.
O’Connor, K. (1997). “Motives and Cognitions in Negotiation: A Theoretical Integration and an Empirical Test,”
International Journal of Conflict Management 8, 114–131.
Olekalns, M. (1991). “The Balance of Power: Effects of Role and Market Forces on Negotiated Outcomes,” Journal
of Applied Social Psychology 21, 1012–1033.
Pfeffer, J. (1992). Managing with Power, Boston, MA: Harvard Business School Press.
Pinkley, R. L. (1995). “Impact of Knowledge Regarding Alternatives to Settlement in Dyadic Negotiation: Whose
Knowledge Counts?” Journal of Applied Psychology 80, 403–417.
Pinkley, R. L., M. A. Neale, and R. J. Bennett. (1994). “The Impact of Alternatives to Settlement in Dyadic
Negotiation,” Organizational Behavior and Human Decision Processes 57, 97–116.
Pittutla, M. and J. K. Murnighan. (2003). Power Lost, Power Gained: Egocentric Action and Inconsistent Perceptions, Working Paper, Kellogg school of Management, Northwestern University.
Raiffa, H. (1982). The Art and Science of Negotiation, Cambridge, MA: Belknap/Harvard University Press.
Ritov, I. (1996). “Anchoring in Simulated Competitive Market Negotiation,” Organizational Behavior and Human
Decision Processes 66, 228–237.
Ross, L. and R. E. Nisbett. (1991). The Person and the Situation, Boston: McGraw-Hill.
Rubin, J. Z. and B. R. Brown. (1975). The Social Psychology of Bargaining and Negotiation, San Diego, CA:
Academic Press.
Rubin, J. Z., D. G. Pruitt, and S. H. Kim. (1994). Social Conflict: Escalation. Stalemate and Settlment, Boston:
McGraw-Hill.
Salancik, G. R. and J. Pfeffer. (1977). “Who Gets Power and How They Hold on to it: A Strategic Contingency
Model of Power,” Organizational Dynamics 5, 3–25.
Sondak, H. and M. H. Bazerman. (1991). “Power Balance and the Rationality of Outcomes in Matching Markets,”
Organizational Behavior and Human Decision Processes 50, 1–23.
Snodgrass, S. E., M. A. Hecht, and R. Ploutz-Snyder. (1998). “Interpersonal Sensitivity: Expressivity or Perceptivity,” Journal of Personality and Social Psychology 74, 238–249.
Thibaut, J. W. and H. H. Kelley. (1959). The Social Psychology of Groups, New York: Wiley.
Thompson, L. and T. DeHarpport. (1994). “Social Judgment, Feedback, and Interpersonal Learning in Negotiation,”
Organizational Behavior and Human Decision Processes 58, 327–345.
Weingart, L. R., L. L. Thompson, M. H. Bazerman, and J. S. Carroll. (1990). “Tactical Behavior and Negotiation
Outcomes,” The International Journal of Conflict Management 1, 7–31.
Winter, D. G. (1973). The Power Motive, New York: Free Press.
Zartman, I. W. and J. Z. Rubin (eds.). (2000). Power and Negotiation, Ann Arbor: University of Michigan Press.
Zetik, D. C. and A. F. Sthuhlmacher. (2002). “Goal Setting and Negotiation Performance: A Meta-Analysis,”
Group Processes and Intergroup Relations 5, 35–52.